PreciseMail Case Study: Middlesex University

 

PreciseMail Case Study: Middlesex University (PDF)

Download a PDF version of this document

 

Middlesex University Chooses PreciseMail Anti-Spam Gateway– 99% of Spam Eliminated for Over 24,000 Email Users

 

“PreciseMail Anti-Spam Gateway stops 99% of the spam consistently without filtering out legitimate email. After setting up my allowlist, I am getting practically zero false positives. Our users are also experiencing the same results.”

 

Middlesex University Profile

Middlesex University is one of the largest universities in the United Kingdom. It has over 22,000 students and 2,500 staff. There are several campuses across North London. Middlesex University also teaches more international students (over 5,000) than any other United Kingdom university. Offering a wide and flexible range of undergraduate and postgraduate programs, it is also the world leader in Work Based Learning and has major contracts with the NHS and the TTA to train nurses, midwives and teachers.

 

The Situation

Email is an important communications tool at Middlesex University. It is used for administrative purposes and for communication between staff and students. Recent increases in the amount of spam being received had made some users' email accounts almost unusable. "Reviewing every email is time-consuming, but necessary when you are looking at the subject line and can't determine if messages are spam or legitimate," said David Webb, Security Team Leader of CCSS at Middlesex University. "The quantity of spam a user receives daily varies. I am inundated with approximately 100 to 150 spam messages per day."

The University subscribes to three different anti-spam blacklist services, which block known spam sources, registered dialup addresses, open-relays, and more. "These services stop a large amount of spam and were sufficient for many years. However, the growth in spam means that these services alone were no longer sufficient. Spam still accounted for approximately 50% of email coming into the University," said Webb.

Before Webb began to evaluate anti-spam solutions, he identified several key requirements:

  • Accuracy - High spam detection rate with a minimum amount of false positives (loss of legitimate email).
  • Flexibility - The ability to customize a spam filtering solution to specific University requirements.
  • Ease of management and deployment - It was important to deploy a solution centrally on the University's mailhubs. Webb did not want to have to purchase, deploy and support a myriad of solutions associated with different email clients and web browsers (allowing web based access to email) on the desktop.
  • User control over spam definitions - The final disposition of messages needed to be under user control. Middlesex University did not want to be in the position of censoring/deleting email just because an anti-spam product thinks that a particular message is spam.

Webb evaluated anti-spam freeware software and several solutions offered by anti-virus vendors. He rejected the freeware as a viable solution for the University because it was not easy to deploy, had poor performance, and lacked the flexibility of a web-based interface.

The anti-spam products offered by anti-virus vendors treated spam in a similar way to viruses. "The problem is that whereas false positives for viruses are extremely low, spam is much less well defined. One person's spam is another person's important email message," said Webb. The anti-virus vendors provided only minimal user-based controls over spam definitions.

 

“The web-based interface is self-explanatory. The intuitive interface combined with user education resulted in few Help Desk inquiries.”

 

The Solution

After considerable testing and feedback from users, Middlesex University implemented PreciseMail Anti-Spam Gateway software as their solution. PreciseMail Anti-Spam Gateway was easy to install on the University's mailhubs. "It stops 99% of the spam consistently without filtering out legitimate email. After setting up my allowlist, I am getting practically zero false positives. Our users are also experiencing the same results," said Webb.

PreciseMail Anti-Spam Gateway's flexibility gave Middlesex University the ability to customize the product to meet their specific requirements. "Although PreciseMail Anti-Spam Gateway is an effective out-of-the-box solution, a key feature is the ability to add and tune filters," said Webb. There are several accounts at the University where students send inquiries about courses from all over the world. Since these accounts are well publicized, they garner a lot of spam. They also have the potential to generate false positives. Creating an allowlist (senders or domains that always bypass the filters) will not solve the problem of false positives because there is no way to predict the email address of the sender ahead of time. As a result, Webb adjusted filtering scores and added some rules in order to adapt the filtering to the type of email the University receives.

Also important to Middlesex University is PreciseMail AntiSpam Gateway's ability to be easily deployed in their existing email environment. "The installation and configuration process on the Alpha and both nodes of the Internal mailhub cluster was fairly simple," said Webb. The University authenticates users’ access to the web-based interface, in order to manage their spam, in several ways. Users whose email is stored on the Alpha cluster are authenticated against the Alpha's IMAP server. Users whose email is stored on one of the internal systems are authenticated against both our Novell NDS (via a POP server) and also against Microsoft's Active Directory (via LDAP).

PreciseMail Anti-Spam Gateway offers the most user-based controls compared to other anti-spam solutions that Webb researched. "Since what is and is not spam is such a slippery concept, users needed to be able to set their own spam filtering definitions," said Webb. PreciseMail Anti-Spam Gateway is configured to tag email that is likely to be spam by default. Users can setup filtering in their email client to move tagged messages into a separate folder for review.

It is then up to the individual user to alter these default settings via the web-based interface by adding addresses into their personal allowlist or blocklists. Users can access the web-based interface to quarantine suspect email on the mailhubs rather than having it tagged and delivered. This email can also be reviewed, deleted, or released through the same web interface. After a certain period, which is configurable by a system administrator, quarantined email is automatically deleted. By default email is deleted after 14 days. However at Middlesex University, quarantined email is set to be deleted automatically after a five-day period. This period will probably be extended in the future once it has been established how many people wish to use quarantining and the amount of disk space this might consume.

 

“One user sent me an email claiming to be ecstatic to have a product that filters the spam so effectively.”

 

Results

Webb is pleased with PreciseMail Anti-Spam Gateway's performance and met his requirement of integrating with the University's mailhubs. Webb reported, "Although PreciseMail Anti-Spam Gateway is running on the same system as the mailhubs, its performance has not been impacted."

Webb measures PreciseMail Anti-Spam Gateway's success by the fact that the Help Desk is no longer getting complaints about spam. They also did not get as many questions from users as they first anticipated. Webb added, "The web-based interface is self-explanatory. The intuitive interface combined with user education resulted in few Help Desk inquiries." In addition, he has received positive feedback from users. "One user sent me an email claiming to be ecstatic to have a product that filters the spam so effectively."